Most cancers grow progressively and evade immunity, despite expression of numerous antigens. Recent evidence has suggested two broad categories of immune escape - a major subset of tumors lacks evidence for T cell-based inflammation and likely represents a situation of immunologic ignorance; but a second subset does indeed contain activated CD8+ T cells that seem to be suppressed through immune inhibitory pathways. The second subset appears to include the majority of patients who respond to currently studied immunotherapies in the clinic, which makes understanding the mechanism driving this spontaneous smoldering anti-tumor immune response in a subset of patients a major fundamental question. Beginning with gene expression profiling of human melanoma samples and through mechanistic studies in mice, we recently have uncovered a critical role for host type I IFN production as a bridge between innate immune sensing and spontaneous CD8+ T cell priming in response to a growing tumor in vivo. But what host innate immune pathway detects a tumor product and drives type I IFN production? Unexpectedly, our preliminary data have suggested that it is tumor-derived DNA which activates a pathway involving the adapter STING to drive IFN-? production by APCs. In STING-/- mice, spontaneous T cell priming by tumors is ablated, and immunogenic tumors that normally are spontaneously rejected grow progressively. The overall goal of this proposal is to understand the mechanism by which the host STING pathway becomes activated by tumors in vivo, and to identify the level of block in this pathway in non-inflamed human melanomas. In the first specific aim, we will investigate the mechanisms by which DNA becomes transferred to host APCs to drive IFN-? production in response to tumor challenge in vivo. Multiple methods will be used to track tumor derived DNA transfer and to phenotype the APCs taking up DNA and responding with IFN-? production. The role for host DNase II in limiting this response will be determined. A novel inducible genetic melanoma model will be used to supplement data with transplantable tumors. In the second specific aim, the regulation of the host STING pathway by candidate DNA sensors and regulators will be evaluated. The focus will be on cGAS, p204, and AIM2. Impact on signaling through the STING pathway will be studied in vitro, and anti-tumor immunity will be interrogated in vivo using gene-targeted mice. In the third specific aim, human melanoma samples already characterized for the level of CD8+ T cell infiltration will be evaluated for the level of block inthis tumor DNA sensing pathway to decipher the reason for failed spontaneous immune recognition in a major subset of patients. Together, these results will elucidate a new fundamental process involved in the host recognition of tumors, and will enable the design of new strategies to promote effective anti-tumor immunity in vivo, particularly in cases that lack spontaneous host immune responses.